Modulation mode recognition method and device based on comprehensive characteristics of wireless signals

Abstract

The application belongs to the technical field of signal processing, and provides a modulation mode recognition method and device based on wireless signal comprehensive features, which comprises the following steps: obtaining the comprehensive features of a target from a wireless signal, wherein the comprehensive features comprise at least one of time domain features, frequency domain features and time-frequency domain features; in the case that the comprehensive features satisfy the feature decision threshold corresponding to the target modulation type and the signal-to-noise ratio threshold, determining the target modulation type as the modulation mode recognition result; wherein the candidate modulation types comprise at least one of 2ASK, 2 / 4 / 8FSK, MSK, 2 / 4 / 8PSK, OQPSK, pi / 4DQPSK and 16QAM. The method improves the modulation mode recognition accuracy of the wireless signal.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of signal processing, and in particular to a modulation mode recognition method and device based on comprehensive features of wireless signals. BACKGROUND

[0002] With the rapid development of modern communication technology, wireless communication systems are widely used in military, civilian and industrial control fields; since the modulation mode directly determines the transmission characteristics of the signal, the accurate recognition of the modulation mode of unknown signals has become an important research topic in the fields of wireless signal monitoring, spectrum supervision and non-cooperative communication.

[0003] In practical applications, the modulation mode of wireless signals includes linear modulation modes (such as AM, FM), nonlinear modulation modes (such as PSK, QAM) and composite modulation modes (such as MSK+AM, 2FSK+FM) and the like, and with the development of high-order modulation technology, the spectrum utilization of signals is significantly improved, but at the same time, it also brings higher challenges to the recognition of modulation modes under low signal-to-noise ratio threshold, complex interference and unknown environment.

[0004] In related technologies, the signal modulation mode is mainly identified and extracted by using a single signal feature, and the identification condition is only related to the signal feature, which is difficult to maintain high recognition accuracy in complex environments (different signal-to-noise ratio threshold requirements), resulting in a decrease in the accuracy of the recognition of the modulation mode of wireless signals. SUMMARY

[0005] The present application provides a modulation mode recognition method and device based on comprehensive features of wireless signals, which solves the defect that the prior art uses a single signal feature to identify and extract the signal modulation mode, and the identification condition is only related to the signal feature, which is difficult to maintain high recognition accuracy in complex environments, resulting in a decrease in the accuracy of the recognition of the modulation mode of wireless signals, and improves the accuracy of the recognition of the modulation mode of wireless signals.

[0006] The present application provides a modulation mode recognition method based on comprehensive features of wireless signals, comprising:

[0007] Obtaining comprehensive features of a target from a wireless signal, the comprehensive features including at least one of time domain features, frequency domain features and time-frequency domain features;

[0008] identify the wireless signal based on a candidate modulation type, and determine the target modulation type as the modulation mode recognition result in a case where the comprehensive feature meets a feature decision threshold corresponding to the target modulation type and a signal-to-noise ratio threshold; wherein the candidate modulation type comprises at least one of 2ASK, 2 / 4 / 8FSK, MSK, 2 / 4 / 8PSK, OQPSK, π / 4DQPSK and 16QAM; and the target modulation type belongs to the candidate modulation type.

[0009] According to the modulation mode recognition method based on the comprehensive feature of the wireless signal provided in the present application, the time domain feature comprises envelope smoothness, and the envelope smoothness is used to distinguish amplitude modulation signals, phase modulation signals and frequency modulation signals.

[0010] The frequency domain feature comprises a peak value number and a flatness of a normalized power spectrum, and the frequency domain feature is used to represent frequency characteristics of the wireless signal.

[0011] The time-frequency domain feature comprises a flatness of a modulus of a normalized instantaneous frequency, and the time-frequency domain feature is used to identify frequency variation characteristics of the wireless signal.

[0012] According to the modulation mode recognition method based on the comprehensive feature of the wireless signal provided in the present application, the comprehensive feature comprises at least two of the time domain feature, the frequency domain feature and the time-frequency domain feature.

[0013] The identifying the wireless signal based on the candidate modulation type comprises:

[0014] The wireless signal is identified based on the candidate modulation type in a target order, and the target modulation type first appearing is determined as the modulation mode recognition result in a case where the comprehensive feature meets a feature decision threshold corresponding to the target modulation type and a signal-to-noise ratio threshold; wherein the target order is determined based on the time domain feature, the frequency domain feature and the time-frequency domain feature.

[0015] According to the modulation mode recognition method based on the comprehensive feature of the wireless signal provided in the present application, the comprehensive feature comprises a time domain feature; and the time domain feature comprises envelope smoothness.

[0016] The determining the target modulation type as the modulation mode recognition result in a case where the comprehensive feature meets a feature decision threshold corresponding to the target modulation type and a signal-to-noise ratio threshold comprises:

[0017] If the feature decision threshold corresponding to the time-domain feature is greater than 1.27 and the signal-to-noise ratio threshold corresponding to the time-domain feature is greater than or equal to 5dB, the modulation scheme identification result is determined to be 2ASK.

[0018] If the feature decision threshold corresponding to the time-domain feature is between 1.12 and 1.27, and the signal-to-noise ratio threshold corresponding to the time-domain feature is greater than or equal to 12dB, the modulation scheme identification result is determined to be 2PSK or 16QAM.

[0019] If the feature decision threshold corresponding to the time-domain feature is between 1.03 and 1.12, and the signal-to-noise ratio threshold corresponding to the time-domain feature is greater than or equal to 12dB, the modulation scheme identification result is determined to be 4PSK, 8PSK, π / 4DQPSK or 4OQPSK.

[0020] If the feature decision threshold corresponding to the time-domain feature is less than 1.03 and the signal-to-noise ratio threshold corresponding to the time-domain feature is greater than or equal to 15dB, the modulation scheme identification result is determined to be CW, 2MSK or 2 / 4 / 8FSK.

[0021] Alternatively, if the feature decision threshold corresponding to the time-domain feature is less than 1.03 and the signal-to-noise ratio threshold corresponding to the time-domain feature is greater than or equal to 12dB, the modulation scheme identification result is determined to be 2MSK or 2 / 4 / 8FSK.

[0022] According to the modulation scheme identification method based on the comprehensive features of wireless signals provided by the present invention, the comprehensive features further include frequency domain features; the frequency domain features include the number of peaks in the normalized power spectrum;

[0023] The step of determining the target modulation type as a modulation scheme identification result when the comprehensive features satisfy the feature decision threshold and signal-to-noise ratio threshold corresponding to the target modulation type further includes:

[0024] If the power spectra of the wireless signal at the 1st, 2nd, 4th, and 8th powers converge and the number of peaks is 1, the modulation scheme identification result is determined to be CW.

[0025] When the power spectra of the wireless signal at the 1st, 2nd, 4th and 8th powers converge, the power spectrum at the 4th power corresponds to a signal-to-noise ratio threshold greater than or equal to 0dB, the power spectrum at the 8th power corresponds to a signal-to-noise ratio threshold greater than or equal to 7dB, and the number of peaks is 1, the modulation method identification result is determined to be 2ASK.

[0026] determining that the modulation mode recognition result is 2PSK in a case where the power spectrum of the order of 2, 4 and 8 of the wireless signal converges, the power spectrum of the order of 4 corresponds to a signal-to-noise ratio threshold greater than or equal to 1dB, the power spectrum of the order of 8 corresponds to a signal-to-noise ratio threshold greater than or equal to 8.5dB, and the number of peaks is all 1;

[0027] determining that the modulation mode recognition result is 4PSK or 4OQPSK in a case where the power spectrum of the order of 4 of the wireless signal converges, the power spectrum of the order of 8 partially converges, the power spectrum of the order of 4 corresponds to a signal-to-noise ratio threshold greater than or equal to 4dB, and the number of peaks is all 1;

[0028] determining that the modulation mode recognition result is 2MSK in a case where the power spectrum of the order of 2, 4 and 8 of the wireless signal converges, the power spectrum of the order of 2 corresponds to a signal-to-noise ratio threshold greater than or equal to 1dB, the power spectrum of the order of 4 corresponds to a signal-to-noise ratio threshold greater than or equal to 7dB, the power spectrum of the order of 8 corresponds to a signal-to-noise ratio threshold greater than or equal to 13.5dB, and the number of peaks is all 2.

[0029] According to the application, a modulation mode recognition method based on comprehensive features of a wireless signal is provided, wherein the frequency domain features further include a flatness degree.

[0030] The determining of the target modulation type as the modulation mode recognition result in a case where the comprehensive features satisfy the feature decision threshold and the signal-to-noise ratio threshold corresponding to the target modulation type further includes:

[0031] determining that the modulation mode recognition result is CW in a case where the feature decision threshold corresponding to the power spectrum of the order of 1 of the wireless signal is greater than 0.3, the signal-to-noise ratio threshold is greater than -6dB, the feature decision threshold corresponding to the power spectrum of the order of 2 is greater than 0.32, the signal-to-noise ratio threshold is greater than -6dB, and the feature decision threshold corresponding to the power spectrum of the order of 4 is greater than 0.3, the signal-to-noise ratio threshold is greater than -6dB.

[0032] determining that the modulation mode recognition result is 2ASK in a case where the feature decision threshold corresponding to the power spectrum of the order of 1 of the wireless signal is greater than 0.3, the signal-to-noise ratio threshold is greater than -6dB, the feature decision threshold corresponding to the power spectrum of the order of 2 is greater than 0.32, the signal-to-noise ratio threshold is greater than 1dB, and the feature decision threshold corresponding to the power spectrum of the order of 4 is greater than 0.27, the signal-to-noise ratio threshold is greater than 3dB.

[0033] In the case that the power spectrum of the 1st power of the wireless signal corresponds to a feature decision threshold less than 0.3, a signal-to-noise ratio threshold greater than -6dB, the power spectrum of the 2nd power corresponds to a feature decision threshold greater than 0.32, a signal-to-noise ratio threshold greater than 1dB, and the power spectrum of the 4th power corresponds to a feature decision threshold greater than 0.27, a signal-to-noise ratio threshold greater than 3dB, the modulation mode recognition result is determined as 2PSK.

[0034] In the case that the power spectrum of the 1st power of the wireless signal corresponds to a feature decision threshold less than 0.3, a signal-to-noise ratio threshold greater than -6dB, the power spectrum of the 2nd power corresponds to a feature decision threshold less than 0.1, a signal-to-noise ratio threshold greater than or equal to -6dB, and the power spectrum of the 4th power satisfies one of the following conditions:

[0035] In the case that the feature decision threshold of the power spectrum of the 4th power is between 0.1 and 0.4, and the signal-to-noise ratio threshold is greater than 3dB, the modulation mode recognition result is determined as 4PSK or π / 4DQPSK.

[0036] In the case that the feature decision threshold of the power spectrum of the 4th power is between 0.25 and 0.4, and the signal-to-noise ratio threshold is greater than 6dB, the modulation mode recognition result is determined as 4OQPSK.

[0037] In the case that the feature decision threshold of the power spectrum of the 4th power is less than 0.05, and the signal-to-noise ratio threshold is greater than or equal to -6dB, the modulation mode recognition result is determined as 8PSK.

[0038] In the case that the power spectrum of the 1st power of the wireless signal corresponds to a feature decision threshold less than 0.3, a signal-to-noise ratio threshold greater than -6dB, the power spectrum of the 2nd power corresponds to a feature decision threshold less than 0.3, a signal-to-noise ratio threshold greater than or equal to -3dB, and the power spectrum of the 4th power satisfies one of the following conditions:

[0039] In the case that the feature decision threshold of the power spectrum of the 4th power is between 0.15 and 0.35, and the signal-to-noise ratio threshold is greater than 12dB, the modulation mode recognition result is determined as 2MSK.

[0040] In the case that the feature decision threshold of the power spectrum of the 4th power is less than 0.05, and the signal-to-noise ratio threshold is greater than -6dB, the modulation mode recognition result is determined as 8FSK.

[0041] According to the present application, a modulation mode recognition method based on the comprehensive features of a wireless signal is provided, wherein the comprehensive features further include time-frequency domain features; the time-frequency domain features include the flatness of the modulus of the normalized instantaneous frequency.

[0042] The determining the target modulation type as the modulation mode recognition result in the case that the comprehensive feature meets a feature decision threshold corresponding to the target modulation type and a signal-to-noise ratio threshold further includes:

[0043] In the case that the feature decision threshold corresponding to the flatness degree of the mode is between 1.8 and 2.5, and the corresponding signal-to-noise ratio threshold is greater than or equal to 10 dB, the modulation mode recognition result is determined as π / 4DQPSK.

[0044] In the case that the feature decision threshold corresponding to the flatness degree of the mode is greater than 1.9, and the corresponding signal-to-noise ratio threshold is greater than or equal to 8 dB, the modulation mode recognition result is determined as 2ASK, 2 / 4 / 8PSK or 16QAM.

[0045] In the case that the feature decision threshold corresponding to the flatness degree of the mode is between 1.32 and 1.95, and the corresponding signal-to-noise ratio threshold is greater than or equal to 12 dB, the modulation mode recognition result is determined as 4OQPSK or 4 / 8FSK.

[0046] In the case that the feature decision threshold corresponding to the flatness degree of the mode is between 1.05 and 1.32, and the corresponding signal-to-noise ratio threshold is greater than or equal to 20 dB, the modulation mode recognition result is determined as 2FSK or 2MSK.

[0047] In the case that the feature decision threshold corresponding to the flatness degree of the mode is less than 1.05, and the corresponding signal-to-noise ratio threshold is greater than or equal to 13 dB, the modulation mode recognition result is determined as CW.

[0048] The application further provides a modulation mode recognition device based on comprehensive features of wireless signals, characterized by comprising:

[0049] A feature acquisition module is configured to acquire comprehensive features from wireless signals, wherein the comprehensive features include at least one of time domain features, frequency domain features and time-frequency domain features.

[0050] An identification module is configured to identify the wireless signals based on candidate modulation types, and determine a target modulation type as a modulation mode recognition result in the case that the comprehensive features meet a feature decision threshold corresponding to the target modulation type and a signal-to-noise ratio threshold; wherein the candidate modulation types include at least one of 2ASK, 2 / 4 / 8FSK, MSK, 2 / 4 / 8PSK, OQPSK, π / 4DQPSK and 16QAM; and the target modulation type belongs to the candidate modulation types.

[0051] The application further provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the modulation mode identification method of the wireless signal comprehensive feature according to any one of the above when executing the computer program.

[0052] The application further provides a non-transitory computer readable storage medium, which stores a computer program, wherein the computer program is executed by a processor to implement the modulation mode identification method of the wireless signal comprehensive feature according to any one of the above.

[0053] The application further provides a computer program product, comprising a computer program, wherein the computer program is executed by a processor to implement the modulation mode identification method of the wireless signal comprehensive feature according to any one of the above.

[0054] The application provides the modulation mode identification method and device based on the wireless signal comprehensive feature, which obtains the comprehensive feature from the wireless signal, and identifies the wireless signal based on the candidate modulation type, and determines the target modulation type as the modulation mode identification result in the case that the comprehensive feature meets the feature decision threshold and the signal-to-noise ratio threshold corresponding to the target modulation type, thereby improving the modulation mode identification accuracy of the wireless signal. BRIEF DESCRIPTION OF DRAWINGS

[0055] In order to more clearly illustrate the technical solutions of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor.

[0056] Figure 1 Fig. 1 is one of the flow diagrams of the modulation mode identification method based on the wireless signal comprehensive feature provided by the present application.

[0057] Figure 2 Fig. 2 is another of the flow diagrams of the modulation mode identification method based on the wireless signal comprehensive feature provided by the present application.

[0058] Figure 3 Fig. 3 is a structural diagram of the modulation mode identification device based on the wireless signal comprehensive feature provided by the present application.

[0059] Figure 4 Fig. 4 is a structural diagram of the electronic device provided by the present application. DETAILED DESCRIPTION

[0060] In order to make the objects, technical solutions and advantages of the present application clearer, the technical solutions in the present application will be described clearly and completely below with reference to the drawings in the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0061] The present application provides a modulation mode recognition method and device based on comprehensive features of wireless signals. Figures 1-3 The present application provides a modulation mode recognition method and device based on comprehensive features of wireless signals.

[0062] Figure 1 The present application provides a modulation mode recognition method and device based on comprehensive features of wireless signals. Figure 1 The present application provides a modulation mode recognition method and device based on comprehensive features of wireless signals.

[0063] In step 110, the comprehensive features of the target are obtained from the wireless signals, and the comprehensive features include at least one of time domain features, frequency domain features and time-frequency domain features.

[0064] In this step, the target in the wireless signals includes time domain baseband in-phase and quadrature (IQ) data of the wireless signals.

[0065] In this step, the time domain features include changes of amplitudes and phases, such as high-order cumulants, statistical features of instantaneous amplitudes, etc.

[0066] In this embodiment, the time domain features include a flatness of a signal envelope; the signal envelope can stably reflect different modulation signal features, and the flatness of the signal envelope can be used to identify the modulation mode. iqR In this step, the time domain features include a flatness of a signal envelope; the signal envelope can stably reflect different modulation signal features, and the flatness of the signal envelope can be used to identify the modulation mode.

[0067] In this embodiment, for 2ASK, 2FSK, 4FSK, 8FSK, MSK, 2PSK, 4PSK, 8PSK, OQPSK, π / 4DQPSK and 16QAM, the value of the flatness of the signal envelope of the 2ASK and 16QAM with amplitude modulation should be greater than that of the 2PSK, 4PSK, 8PSK, OQPSK and π / 4DQPSK with phase modulation, and the value of the flatness of the signal envelope of the 2FSK, 4FSK, 8FSK and MSK with frequency modulation should be less than that of the above-mentioned modulation signals. iqR In this embodiment, for 2ASK, 2FSK, 4FSK, 8FSK, MSK, 2PSK, 4PSK, 8PSK, OQPSK, π / 4DQPSK and 16QAM, the value of the flatness of the signal envelope of the 2ASK and 16QAM with amplitude modulation should be greater than that of the 2PSK, 4PSK, 8PSK, OQPSK and π / 4DQPSK with phase modulation, and the value of the flatness of the signal envelope of the 2FSK, 4FSK, 8FSK and MSK with frequency modulation should be less than that of the above-mentioned modulation signals. iqR In this embodiment, for 2ASK, 2FSK, 4FSK, 8FSK, MSK, 2PSK, 4PSK, 8PSK, OQPSK, π / 4DQPSK and 16QAM, the value of the flatness of the signal envelope of the 2ASK and 16QAM with amplitude modulation should be greater than that of the 2PSK, 4PSK, 8PSK, OQPSK and π / 4DQPSK with phase modulation, and the value of the flatness of the signal envelope of the 2FSK, 4FSK, 8FSK and MSK with frequency modulation should be less than that of the above-mentioned modulation signals. iqR In this embodiment, for 2ASK, 2FSK, 4FSK, 8FSK, MSK, 2PSK, 4PSK, 8PSK, OQPSK, π / 4DQPSK and 16QAM, the value of the flatness of the signal envelope of the 2ASK and 16QAM with amplitude modulation should be greater than that of the 2PSK, 4PSK, 8PSK, OQPSK and π / 4DQPSK with phase modulation, and the value of the flatness of the signal envelope of the 2FSK, 4FSK, 8FSK and MSK with frequency modulation should be less than that of the above-mentioned modulation signals.

[0068] In this embodiment, the frequency domain features include a peak value number of normalized power spectrums of different powers (for example, 1, 2, 4 or 8 powers) of the wireless signals and a flatness of the normalized power spectrums.

[0069] Specifically, the peak number of the normalized power spectrum is obtained in the following manner: first, the wireless signal is raised to the power of 1, 2, 4, and 8, respectively, and then a linear power spectrum is calculated, and the maximum value is taken as a reference to normalize the linear power spectrum, and the peak number of the normalized power spectrum is calculated.

[0070] For 2ASK, 2 / 4 / 8FSK, MSK, 2 / 4 / 8PSK, OQPSK, π / 4DQPSK, and 16QAM, the peak number of the normalized power spectrum includes 1, 2, 1+2 (one main peak and two sideband symbol rate spectral lines), 4, and 8 in these five cases. In this embodiment, the positions of possible peaks are first found out, and the case of non-single peak caused by spectrum leakage (the points near the main peak are close to the main peak) is also considered.

[0071] In this embodiment, the flatness kur of the normalized power spectrum corresponding to the signal raised to the power of 1, 2, 4, and 8 is calculated.

[0072] Specifically, the maximum value of the linear power spectrum of the signal is normalized, and then the normalized power spectrum is smoothed, and then the ratio of the eighth central moment of the power spectrum to the square of the fourth origin moment of the power spectrum is calculated, and then the length of the signal is divided to obtain the above-mentioned kur.

[0073] In this embodiment, the time-frequency domain feature includes the flatness of the modulus of the normalized instantaneous frequency of the signal. The instantaneous frequency represents the relationship between time and frequency, which can be obtained by phase difference. Then the instantaneous frequency is normalized, for example, the instantaneous frequency is mapped to the interval [0, 1] to obtain the normalized instantaneous frequency. Then the modulus of the normalized instantaneous frequency is calculated, and the flatness thereof is evaluated to obtain the flatness of the modulus of the normalized instantaneous frequency. .

[0074] Step 120, based on the candidate modulation type, the wireless signal is identified, and in the case that the comprehensive feature meets the feature decision threshold and the signal-to-noise ratio threshold corresponding to the target modulation type, the target modulation type is determined as the modulation mode recognition result; wherein the candidate modulation type includes at least one of 2ASK, 2 / 4 / 8FSK, MSK, 2 / 4 / 8PSK, OQPSK, π / 4DQPSK, and 16QAM; the target modulation type belongs to the candidate modulation type.

[0075] In this step, the feature decision threshold and the signal-to-noise ratio threshold can be obtained through simulation experiment; for example, using an awgn channel as a simulated channel model to record the feature decision threshold and the signal-to-noise ratio threshold corresponding to the comprehensive feature, to obtain the corresponding decision interval; for example, the comprehensive feature includes the flatness of the signal envelope iqR , which is determined by simulation:

[0076] WheniqR When the corresponding feature decision threshold is greater than 1.27 and the required signal-to-noise ratio threshold SNR is greater than or equal to 5 dB, the modulation mode recognition result of the wireless signal is determined as 2ASK; iqR Corresponding to different feature decision thresholds and signal-to-noise ratio thresholds, the wireless signal corresponds to different modulation modes.

[0077] For example, the comprehensive feature includes a frequency domain feature, and the frequency domain feature includes a peak number peakNum of a normalized power spectrum. Through simulation, it is determined that the 2th, 4th and 8th power spectra of the wireless signal all converge, the signal-to-noise ratio of the 2th power spectrum has no effect on the signal, the SNR of the 4th power spectrum is greater than or equal to 0 dB, the SNR of the 8th power spectrum is greater than or equal to 7 dB, and the peak number of each power spectrum is always 1. Therefore, the modulation mode recognition result of the wireless signal is determined as 2PSK. The peakNum corresponds to different feature decision thresholds and signal-to-noise ratio thresholds, and the wireless signal corresponds to different modulation modes.

[0078] In this embodiment, the frequency domain feature further includes a flatness kur of the normalized power spectrum. Through simulation, it is determined that the kur of the 1st power spectrum of the wireless signal is greater than 0.3, and the signal-to-noise ratio threshold is greater than or equal to -6 dB, the kur of the 2th power spectrum is greater than 0.32, and the signal-to-noise ratio threshold is greater than or equal to -6 dB, the kur of the 4th power spectrum is greater than 0.3, and the signal-to-noise ratio threshold is greater than or equal to -6 dB. Therefore, the modulation mode recognition result of the wireless signal is determined as CW. The kur corresponds to different feature decision thresholds and signal-to-noise ratio thresholds, and the wireless signal corresponds to different modulation modes.

[0079] For example, the comprehensive feature includes a time-frequency domain feature, and the time domain feature is a flatness ftNormAbsR of a normalized instantaneous frequency. Through simulation, it is determined that the decision interval of the ftNormAbsR corresponds to between 1.8 and 2.5, and the required signal-to-noise ratio SNR is greater than or equal to 10 dB. Therefore, the modulation mode recognition result of the wireless signal is determined as π / 4DQPSK. The ftNormAbsR corresponds to different feature decision thresholds and signal-to-noise ratio thresholds, and the wireless signal corresponds to different modulation modes.

[0080] In this embodiment, for the input unknown signal, the comprehensive feature is extracted first, and each candidate modulation type (2ASK, 2FSK, 4FSK, 8FSK, MSK, 2PSK, 4PSK, 8PSK, OQPSK, π / 4DQPSK, 16QAM) is traversed. If the parameter value corresponding to the comprehensive feature meets the decision threshold and the signal-to-noise ratio threshold of the corresponding feature of the current candidate modulation type at the same time when a certain modulation mode is traversed, the current input signal belongs to the modulation mode this time, and the blind recognition of the modulation mode of the unknown modulation type signal is completed.

[0081] In this embodiment, for the signal processing parameters involved in the modulation mode identification process, the number of symbols contained in the signal to be identified is ≥512, the oversampling multiple is ≥4, and the FFT point number is ≥2048 points. Multiple gears can be set, such as 2048, 4096, and 8192 points, to ensure the identification rate under low signal-to-noise ratio. The identification rates obtained by different classification methods in this embodiment are different, and the above classification method can be continuously adjusted, and the classification threshold can also be fine-tuned.

[0082] In this embodiment, the feature decision threshold and the signal-to-noise ratio threshold corresponding to the comprehensive feature are recorded using the awgn channel as the simulated channel model. The corresponding simulation parameters and results include:

[0083] (1) Simulation conditions; set CW, 2ASK, 2PSK, 4PSK, 8PSK, 4OQPSK, π / 4DQPSK, 16QAM, 2FSK, 4FSK, 8FSK, and 2MSK, a total of 12 kinds of signals (2 / 4 / 8FSK / MSK are signals obtained by linear frequency modulation method), among which CW is used for comparison reference; set the signal-to-noise ratio to -10dB to 35dB, step 0.01dB, symbol rate to 8192Hz, symbol number to 1024 / 512 / 256 / 128, and oversampling multiple to 4.

[0084] (2) Use the above strategy to perform simulation, and the simulation results are shown in Table 1 below:

[0085] Table 1. Classification data table with oversampling multiple of 4 and symbol number of 1024 / 512 / 256 / 128

[0086]

[0087] It should be noted that the percentage in the above table is the number of successful classifications / the number of test classifications × 100%.

[0088] In this embodiment, in order to reflect the randomness of the signal, for each signal-to-noise ratio, the IQ amplitude random value interval is [10, 100], the IQ phase deviation is [0, 2π], the IQ frequency deviation is [-500, 500]Hz, the 2 / 4 / 8FSK modulation index is [0.2, 0.8], the shaping filter roll-off coefficient is [0.2, 0.8] (step 0.1, 2 / 4 / 8FSK is a root raised cosine filter, MSK is a self-contained rectangular pulse, and other signals are randomly generated by a raised cosine filter within the corresponding range), and the message signal is also randomly generated; In this embodiment, the threshold thr when calculating peakNum is set to 0.5, the index difference threshold thrDiff is set to 2, and the smoothing point number L used when calculating kur is 3 points.

[0089] The embodiment of the present application provides a modulation mode recognition method based on comprehensive features of a wireless signal, target comprehensive features are obtained from the wireless signal, and the wireless signal is recognized based on candidate modulation types, in the case that the comprehensive features meet feature decision thresholds and signal-to-noise ratio thresholds corresponding to target modulation types, the target modulation types are determined as modulation mode recognition results, and the modulation mode recognition accuracy of the wireless signal is improved.

[0090] In some embodiments, the time domain feature includes envelope smoothness, and the envelope smoothness is used to distinguish amplitude modulation signals, phase modulation signals and frequency modulation signals.

[0091] In the embodiment, it is assumed that a baseband IQ signal is expressed in a complex number form y, and a signal envelope is The modulus of the complex number y is obtained, and the envelope smoothness is calculated. The variance of the envelope smoothness is calculated. The mean value of the envelope smoothness is calculated. That is, the following equation is obtained:

[0092] ;

[0093] The constant term in the above equation is removed, and the following equation is obtained:

[0094] ;

[0095] wherein the envelope smoothness is a feature parameter of the smoothness degree of the signal envelope, the envelope square mean value is divided by the square of the envelope mean value, iqR The minimum value of the envelope smoothness is 1. iqR The smaller the value of the envelope smoothness is, the more gentle the signal envelope distribution fluctuation is, and the more likely the signal is a non-amplitude modulation signal. R The greater the value of the envelope smoothness is, the more steep the signal envelope distribution fluctuation is, and the more likely the signal is an amplitude modulation signal. iqR The frequency domain feature includes the number and the smoothness degree of the peak value of the normalized power spectrum, and the frequency domain feature is used to represent the frequency characteristics of the wireless signal.

[0096] In the embodiment, the number and the position of the peak value of the normalized power spectrum are obtained by the following steps:

[0097] (1) The value of the normalized power spectrum is first smoothed by L points (linear power spectrum is smoothed), and L can be configured.

[0098]

[0099] ​(2) Set a threshold value, denoted as thr, and compare it with the value of the smoothed normalized power spectrum point by point. When the first point greater than or equal to the threshold appears, it is recorded as the first left index, and the index point is also recorded as the first right index. Then, continue to compare the threshold, and record the last index point that satisfies the condition of being greater than or equal to the threshold until a point that does not satisfy the condition appears. Update the value of the first right index to the value of the last index point that satisfies the condition of being greater than or equal to the threshold. The two index values are the first left and right index pair. Then, continue to repeat the above operation to find the second, third, etc. left and right index pairs until all points are compared. A total of N index pairs are obtained, and the possible number of peak values is N.

[0100] (3) Find the index of the maximum value in each index pair, and set the index difference threshold between the maximum values thrDiff, which is used to eliminate false peak values, including the following cases:

[0101] (4) If N = 1, then the number of peak values peakNum = 1.

[0102] (5) If N = 2, the result of subtracting the first index value from the second index value is greater than thrDiff, then the number of peak values peakNum = 2, otherwise the number of peak values peakNum = 1.

[0103] (6) If N = 3, the peak values corresponding to the middle index value are greater than 0, and the result of subtracting the sum of the front and rear index values from twice the middle index value is less than or equal to thrDiff, then the peak values on both sides are considered to be symbol rate spectral lines, which are false peak values, and the number of peak values peakNum = 1, otherwise the number of peak values peakNum = 3.

[0104] (7) If N = 4 or 8, subtract the first index from the second index to obtain N-1 index differences. Find the mean of the index differences, subtract the mean from each index difference, and take the absolute value to obtain N-1 absolute values. Compare these N-1 absolute values with thrDiff in turn. If the absolute value is less than or equal to thrDiff, then the number of peak values is incremented by 1. If the final number of peak values is N-1, then the difference between the N adjacent peak values is basically consistent, and the number of peak values peakNum = N. If the final number of peak values is not N-1, then the difference between the N adjacent peak values is inconsistent, and the number of peak values peakNum = 0.

[0105] (8) In other cases, the number of peak values peakNum = 0.

[0106] In this embodiment, when calculating the flatness of the normalized power spectrum kur, the linear power spectrum of the signal is first normalized by the maximum value, then smoothed and normalized again, and then the ratio of the 8th central moment of this power spectrum to the square of the 4th origin moment of this power spectrum is calculated, and then divided by the length of the signal to obtain.

[0107] For example, assuming that the power spectrum to be input is , the kth central moment is , the kth origin moment is , that is:

[0108] ;

[0109] , wherein may be the power spectrum after p=1, 2, 4, 8 times of the signal, and len is the length of ; is the mean of the envelope fourth power, and is the square of the envelope mean.

[0110] The time-frequency domain feature includes the flatness of the modulus of the normalized instantaneous frequency, and the time-frequency domain feature is used to identify the frequency variation characteristics of the wireless signal.

[0111] In this embodiment, the time-frequency domain feature is obtained by the following steps: assuming that an IQ complex signal of N points with a sampling rate of fs is received, the arctangent of the signal in the time domain of the N-point signal is calculated to obtain a non-continuous phase sequence ang in the range of [-π, +π], and the difference between the front and rear phases ; wherein the original solution of the instantaneous frequency process involves two times of difference (later minus earlier) operation in the process of unwinding and differentiation, which can be simplified to one operation when applied to engineering, and the instantaneous frequency corresponding to the IQ complex signal is expressed as follows:

[0112] ;

[0113] , wherein ; , and satisfies the following formula:

[0114] ;

[0115] The above is the instantaneous frequency, which is normalized by dividing by 0.5 times the sampling rate and taking the modulus value, and is expressed by the following formula:

[0116] ;

[0117] , wherein is the modulus of the normalized instantaneous frequency of the signal, and its range is 0 to 1; finally, The flatness of the modulus of the normalized instantaneous frequency can be obtained by substituting into the above flatness calculation method, and is denoted as .

[0118] The modulation mode recognition method based on the comprehensive features of the wireless signal provided in the embodiment of the application sets the time domain feature to include envelope smoothness, which is used to distinguish amplitude modulation signals, phase modulation signals and frequency modulation signals; sets the frequency domain feature to include the peak number and the flatness of the normalized power spectrum, which is used to represent the frequency characteristics of the wireless signal, and the combination of the peak number and the flatness of the normalized power spectrum enhances the ability to characterize the frequency characteristics of complex signals; and sets the time-frequency domain feature to include the flatness of the modulus of the normalized instantaneous frequency, which is used to identify the frequency variation characteristics of the wireless signal, and effectively makes up for the deficiency of the traditional time-frequency diagram analysis in dynamic frequency variation recognition.

[0119] In some embodiments, the comprehensive features include at least two of the time domain feature, the frequency domain feature and the time-frequency domain feature; the wireless signal is identified based on the candidate modulation types, and in a case where the comprehensive features meet the feature decision threshold corresponding to the target modulation type and the signal-to-noise ratio threshold, the target modulation type is determined as the modulation mode recognition result, including: the wireless signal is identified based on the candidate modulation types in a target order, and in a case where the comprehensive features meet the feature decision threshold corresponding to the target modulation type and the signal-to-noise ratio threshold, the first target modulation type is determined as the modulation mode recognition result; wherein the target order is determined based on the time domain feature, the frequency domain feature and the time-frequency domain feature.

[0120] In this embodiment, when multiple signal features are used for modulation mode decision, the feature parameters that can distinguish the modulation modes with a lower signal-to-noise ratio are preferentially selected, and the feature parameters that can distinguish the modulation modes with a higher signal-to-noise ratio are secondly selected, that is, the above order can be used to judge the signals that are easy to distinguish first, and then judge the signals that are difficult to distinguish.

[0121] In this embodiment, when the peak number and the flatness of the power spectrum of the multiple powers of the signal are used for modulation mode recognition, the modulation mode of the signal can be judged from the first power, and the result is output immediately after the modulation mode is judged.

[0122] In this embodiment, when the comprehensive features include the time domain feature and the frequency domain feature, the target order is to select the modulation modes in turn according to the time domain feature and the frequency domain feature; and when the comprehensive features include the time domain feature, the frequency domain feature and the time-frequency domain feature, the target order is to select the modulation modes in turn according to the time domain feature, the frequency domain feature and the time-frequency domain feature.

[0123] Specifically, for an unknown wireless signal input, each modulation type 2ASK, 2 / 4 / 8FSK, MSK, 2 / 4 / 8PSK, OQPSK, pi / 4DQPSK and 16QAM is traversed, and whether the corresponding decision threshold and signal-to-noise ratio threshold are met is judged according to the characteristic parameters of the smoothness of the signal envelope extracted from the wireless signal iqR , the number of peak values peakNum of the normalized power spectrum, the characteristic parameter kur of the smoothness of the normalized power spectrum, and the characteristic parameter of the modulus of the normalized instantaneous frequency The four parameters are judged iqR whether the peakNum satisfies the corresponding peak value number of different modulation modes, whether the kur satisfies the corresponding decision threshold and signal-to-noise ratio threshold, and whether the corresponding decision threshold and signal-to-noise ratio threshold are met; if the traversal is performed on a certain modulation mode, iqR , the peakNum, the kur and the four parameters simultaneously satisfy the corresponding characteristic decision threshold of the modulation mode, it is considered that the input signal currently belongs to the modulation mode, and thus the blind identification of the modulation mode of the unknown modulation type signal is completed.

[0124] The modulation mode identification method based on the comprehensive characteristics of the wireless signal provided by the embodiment of the application identifies the wireless signal in a target order according to at least one candidate modulation type, determines the target modulation type appearing for the first time as the modulation mode identification result when the comprehensive characteristics satisfy the characteristic decision threshold and the signal-to-noise ratio threshold corresponding to the target modulation type, finds the target in the modulation type range through the traversal of each modulation type once, avoids repeated complex calculations, optimizes the process of the signal modulation mode identification through the efficient threshold decision and the step-by-step feature extraction strategy, and significantly reduces the calculation amount.

[0125] In some embodiments, the comprehensive characteristics include time domain characteristics; the time domain characteristics include envelope smoothness; and in the case where the comprehensive characteristics satisfy the characteristic decision threshold and the signal-to-noise ratio threshold corresponding to the target modulation type, determining the target modulation type as the modulation mode identification result includes:

[0126] In a case that the feature decision threshold corresponding to the time domain feature is greater than 1.27, and the signal-to-noise ratio threshold corresponding to the time domain feature is greater than or equal to 5 dB, the modulation mode recognition result is determined as 2ASK; in a case that the feature decision threshold corresponding to the time domain feature is between 1.12 and 1.27, and the signal-to-noise ratio threshold corresponding to the time domain feature is greater than or equal to 12 dB, the modulation mode recognition result is determined as 2PSK or 16QAM; in a case that the feature decision threshold corresponding to the time domain feature is between 1.03 and 1.12, and the signal-to-noise ratio threshold corresponding to the time domain feature is greater than or equal to 12 dB, the modulation mode recognition result is determined as 4PSK, 8PSK, π / 4DQPSK or 4OQPSK; in a case that the feature decision threshold corresponding to the time domain feature is less than 1.03, and the signal-to-noise ratio threshold corresponding to the time domain feature is greater than or equal to 15 dB, the modulation mode recognition result is determined as CW, 2MSK or 2 / 4 / 8FSK; or, in a case that the feature decision threshold corresponding to the time domain feature is less than 1.03, and the signal-to-noise ratio threshold corresponding to the time domain feature is greater than or equal to 12 dB, the modulation mode recognition result is determined as 2MSK or 2 / 4 / 8FSK.

[0127] In this embodiment, the decision interval of the envelope smoothness iqR (as described above iqR ) is set as shown in Table 2:

[0128] Table 2. iqR decision interval parameter table

[0129]

[0130] In this embodiment, when the comprehensive feature is the envelope smoothness for identifying different modulation modes, reasonable feature decision thresholds and signal-to-noise ratio thresholds are set to ensure that the modulation mode recognition method can operate efficiently under multiple signal-to-noise ratios.

[0131] In some embodiments, the comprehensive feature further includes a frequency domain feature; the frequency domain feature includes a peak value number of a normalized power spectrum; and in a case that the comprehensive feature satisfies a feature decision threshold and a signal-to-noise ratio threshold corresponding to a target modulation type, the target modulation type is determined as the modulation mode recognition result, further including:

[0132] In the case that the power spectrum of the wireless signal at the 1st, 2nd, 4th and 8th power converges, and the number of peaks is all 1, the modulation mode recognition result is determined as CW; in the case that the power spectrum of the wireless signal at the 1st, 2nd, 4th and 8th power converges, the signal-to-noise ratio threshold corresponding to the 4th power spectrum is greater than or equal to 0dB, the signal-to-noise ratio threshold corresponding to the 8th power spectrum is greater than or equal to 7dB, and the number of peaks is all 1, the modulation mode recognition result is determined as 2ASK; in the case that the power spectrum of the wireless signal at the 2nd, 4th and 8th power converges, the signal-to-noise ratio threshold corresponding to the 4th power spectrum is greater than or equal to 1dB, the signal-to-noise ratio threshold corresponding to the 8th power spectrum is greater than or equal to 8.5dB, and the number of peaks is all 1, the modulation mode recognition result is determined as 2PSK; in the case that the power spectrum of the wireless signal at the 4th power converges, the 8th power spectrum partially converges, the signal-to-noise ratio threshold corresponding to the 4th power spectrum is greater than or equal to 4dB, and the number of peaks is all 1, the modulation mode recognition result is determined as 4PSK or 4OQPSK; in the case that the power spectrum of the wireless signal at the 2nd, 4th and 8th power converges, the signal-to-noise ratio threshold corresponding to the 2nd power spectrum is greater than or equal to 1dB, the signal-to-noise ratio threshold corresponding to the 4th power spectrum is greater than or equal to 7dB, the signal-to-noise ratio threshold corresponding to the 8th power spectrum is greater than or equal to 13.5dB, and the number of peaks is all 2, the modulation mode recognition result is determined as 2MSK.

[0133] In this embodiment, for the number of peaks of the normalized power spectrum peakNum, the following determination rules can be obtained through simulation:

[0134] (1) The power spectrum of the wireless signal at the 1st, 2nd, 4th and 8th power all converges, the signal-to-noise ratio has little effect on it, and the number of peaks is always 1, so the modulation mode recognition result is CW;

[0135] (2) The power spectrum of the wireless signal at the 1st, 2nd, 4th and 8th power all converges, the signal-to-noise ratio has little effect on the 1st and 2nd power, the 4th power SNR is greater than or equal to 0dB, the 8th power SNR is greater than or equal to 7dB, and the number of peaks is always 1, so the modulation mode recognition result is 2ASK;

[0136] (3) The power spectrum of the wireless signal at the 2nd, 4th and 8th power all converges, the signal-to-noise ratio has little effect on the 2nd power, the 4th power SNR is greater than or equal to 1dB, the 8th power SNR is greater than or equal to 8.5dB, and the number of peaks is always 1 after convergence, so the modulation mode recognition result is 2PSK;

[0137] (4) The power spectrum of the wireless signal at the 4th power converges, the 4th power SNR is greater than or equal to 4dB, the 8th power does not completely converge and there is a large probability of misjudgment, high power operation causes the bottom noise to rise and affects the judgment, and the number of peaks is always 1 after convergence, so the modulation mode recognition result is 4PSK;

[0138] (5) The 4th power of the wireless signal converges, the 4th power SNR is greater than or equal to 4dB, the 8th power does not completely converge and there is a large probability of misjudgment, and the number of peaks is always 1 after convergence; the double peak of the 2nd power cannot be effectively judged, the number of symbols is not enough, the peak is not obvious, and the modulation mode recognition result is 4OQPSK.

[0139] (6) The 2nd, 4th and 8th powers of the wireless signal converge, the 2nd power SNR is greater than or equal to 1dB, the 4th power SNR is greater than or equal to 7dB, the 8th power SNR is greater than or equal to 13.5dB, and the number of peaks is always 2 after convergence, and the modulation mode recognition result is 2MSK.

[0140] In the embodiment, when the number of peaks of the normalized power spectrum is used to identify different modulation modes, reasonable peak number judgment thresholds and signal-to-noise ratio thresholds are set to ensure that the modulation mode recognition method can operate efficiently under various signal-to-noise ratios.

[0141] In some embodiments, the frequency domain feature further includes a flatness; and in a case where the comprehensive feature satisfies a feature judgment threshold and a signal-to-noise ratio threshold corresponding to a target modulation type, determining the target modulation type as the modulation mode recognition result further includes:

[0142] In the case that the characteristic decision threshold of the 1st power spectrum of the wireless signal is greater than 0.3, the corresponding signal-to-noise ratio threshold is greater than -6dB, the characteristic decision threshold of the 2nd power spectrum is greater than 0.32, the corresponding signal-to-noise ratio threshold is greater than -6dB, and the characteristic decision threshold of the 4th power spectrum is greater than 0.3, the corresponding signal-to-noise ratio threshold is greater than -6dB, the modulation mode recognition result is determined as CW; in the case that the characteristic decision threshold of the 1st power spectrum of the wireless signal is greater than 0.3, the corresponding signal-to-noise ratio threshold is greater than -6dB, the characteristic decision threshold of the 2nd power spectrum is greater than 0.32, the corresponding signal-to-noise ratio threshold is greater than 1dB, and the characteristic decision threshold of the 4th power spectrum is greater than 0.27, the corresponding signal-to-noise ratio threshold is greater than 3dB, the modulation mode recognition result is determined as 2ASK; in the case that the characteristic decision threshold of the 1st power spectrum of the wireless signal is less than 0.3, the corresponding signal-to-noise ratio threshold is greater than -6dB, the characteristic decision threshold of the 2nd power spectrum is greater than 0.32, the corresponding signal-to-noise ratio threshold is greater than 1dB, the characteristic decision threshold of the 4th power spectrum is greater than 0.27, and the corresponding signal-to-noise ratio threshold is greater than 3dB, the modulation mode recognition result is determined as 2PSK; in the case that the characteristic decision threshold of the 1st power spectrum of the wireless signal is less than 0.3, the corresponding signal-to-noise ratio threshold is greater than -6dB, the characteristic decision threshold of the 2nd power spectrum is less than 0.1, the corresponding signal-to-noise ratio threshold is greater than or equal to -6dB, and the 4th power spectrum satisfies one of the following conditions: in the case that the characteristic decision threshold of the 4th power spectrum is between 0.1 and 0.4, the corresponding signal-to-noise ratio threshold is greater than 3dB, the modulation mode recognition result is determined as 4PSK or π / 4DQPSK; in the case that the characteristic decision threshold of the 4th power spectrum is between 0.25 and 0.4, the corresponding signal-to-noise ratio threshold is greater than 6dB, the modulation mode recognition result is determined as 4OQPSK; in the case that the characteristic decision threshold of the 4th power spectrum is less than 0.05, the corresponding signal-to-noise ratio threshold is greater than or equal to -6dB, the modulation mode recognition result is determined as 8PSK; in the case that the characteristic decision threshold of the 1st power spectrum of the wireless signal is less than 0.3, the corresponding signal-to-noise ratio threshold is greater than -6dB, the characteristic decision threshold of the 2nd power spectrum is less than 0.3, the corresponding signal-to-noise ratio threshold is greater than or equal to -3dB, and the 4th power spectrum satisfies one of the following conditions: in the case that the characteristic decision threshold of the 4th power spectrum is between 0.15 and 0.35, the corresponding signal-to-noise ratio threshold is greater than 12dB, the modulation mode recognition result is determined as 2MSK; in the case that the characteristic decision threshold of the 4th power spectrum is less than 0.05, the corresponding signal-to-noise ratio threshold is greater than -6dB, the modulation mode recognition result is determined as 8FSK.

[0143] In this embodiment, the corresponding decision conditions of the characteristic parameter kur of the flatness of the normalized power spectrum can be obtained through simulation, as shown in Table 3 below.

[0144] The embodiment sets reasonable peak value quantity decision threshold and signal-to-noise ratio threshold when different modulation modes are identified according to the flatness of the normalized power spectrum, and ensures that the modulation mode identification method can operate efficiently under multiple signal-to-noise ratios.

[0145] In some embodiments, the comprehensive feature further includes a time-frequency domain feature; the time-frequency domain feature includes the flatness of the modulus of the normalized instantaneous frequency; and in the case that the comprehensive feature meets the feature decision threshold and the signal-to-noise ratio threshold corresponding to the target modulation type, determining the target modulation type as the modulation mode identification result further includes:

[0146] In the case that the modulus flatness corresponding feature decision threshold is between 1.8 and 2.5, and the corresponding signal-to-noise ratio threshold is greater than or equal to 10dB, the modulation mode identification result is determined as π / 4DQPSK; in the case that the modulus flatness corresponding feature decision threshold is greater than 1.9, and the corresponding signal-to-noise ratio threshold is greater than or equal to 8dB, the modulation mode identification result is determined as 2ASK, 2 / 4 / 8PSK or 16QAM; in the case that the modulus flatness corresponding feature decision threshold is between 1.32 and 1.95, and the corresponding signal-to-noise ratio threshold is greater than or equal to 12dB, the modulation mode identification result is determined as 4OQPSK or 4 / 8FSK; in the case that the modulus flatness corresponding feature decision threshold is between 1.05 and 1.32, and the corresponding signal-to-noise ratio threshold is greater than or equal to 20dB, the modulation mode identification result is determined as 2FSK or 2MSK; and in the case that the modulus flatness corresponding feature decision threshold is less than 1.05, and the corresponding signal-to-noise ratio threshold is greater than or equal to 13dB, the modulation mode identification result is determined as CW.

[0147] In this embodiment, the flatness of the modulus of the normalized instantaneous frequency of the signal ftNormAbsR The following decision conditions are obtained through simulation and are shown in Table 4.

[0148] The embodiment sets reasonable peak value quantity decision threshold and signal-to-noise ratio threshold when different modulation modes are identified according to the flatness of the normalized instantaneous frequency of the normalized power spectrum, and ensures that the modulation mode identification method can operate efficiently under multiple signal-to-noise ratios.

[0149] Figure 2 is a flowchart of a modulation mode identification method based on a wireless signal comprehensive feature provided by the present application, in which Figure 2In the embodiment shown, the modulation mode recognition method based on comprehensive features of wireless signals is also implemented through the following steps: (1) determining a modulation signal of an unknown modulation type of amplitude modulation (ASK type), frequency modulation (FSK, MSK type), phase modulation (PSK, OQPSK, DQPSK type), amplitude and phase modulation (QAM, APSK type); (2) obtaining time domain baseband IQ data of the wireless signal; (3) calculating the envelope of the IQ data, and then calculating the feature parameter iqR of the flatness of the signal envelope; (4) calculating the peak number peakNum of the normalized 1st power spectrum after the IQ data is raised to the power of p=1, 2, 4, 8; (5) calculating the flatness kur of the normalized power spectrum corresponding to the signal raised to the power of p=1, 2, 4, 8, respectively; (6) calculating the instantaneous frequency by performing phase calculation and phase difference and unwrapping on the IQ signal, and then calculating the flatness parameter ftNormAbs of the modulus of the normalized instantaneous frequency; (7) traversing each modulation type; (8) determining whether the four parameters iqR, peakNum, kur, and ftNormAbs simultaneously satisfy the decision threshold of the corresponding features of the modulation mode; if yes, outputting the corresponding modulation type as the recognition result (i.e., the modulation mode recognition result); if not, continuing to traverse the next modulation type, and if all the modulation types are traversed, considering that the recognition fails.

[0150] Table 3. kur decision condition data table

[0151]

[0152] Table 4. ftNormAbsR decision interval data table

[0153]

[0154] The modulation mode recognition device based on comprehensive features of wireless signals provided by the present application is described below, and the modulation mode recognition device based on comprehensive features of wireless signals described below can be correspondingly referred to the modulation mode recognition method based on comprehensive features of wireless signals described above.

[0155] Figure 3 FIG. 1 is a structural schematic diagram of the modulation mode recognition device based on comprehensive features of wireless signals provided by the present application, as shown in the figure, the modulation mode recognition device based on comprehensive features of wireless signals comprises a feature acquisition module 310 and a recognition module 320. Figure 3

[0156] The feature acquisition module 310 is configured to acquire the comprehensive features of the target from the wireless signal, and the comprehensive features include at least one of the time domain features, the frequency domain features, and the time-frequency domain features.

[0157] ​The identification module 320 is configured to identify the wireless signal based on the candidate modulation type, and determine the target modulation type as the modulation mode identification result in a case where the comprehensive feature meets a feature decision threshold corresponding to the target modulation type and a signal-to-noise ratio threshold; the candidate modulation type includes at least one of 2ASK, 2 / 4 / 8FSK, MSK, 2 / 4 / 8PSK, OQPSK, pi / 4DQPSK and 16QAM; and the target modulation type belongs to the candidate modulation type.

[0158] The modulation mode identification device based on the comprehensive feature of the wireless signal provided by the embodiment of the application can obtain the comprehensive feature of the target from the wireless signal, identify the wireless signal based on the candidate modulation type, and determine the target modulation type as the modulation mode identification result in a case where the comprehensive feature meets a feature decision threshold corresponding to the target modulation type and a signal-to-noise ratio threshold, thereby improving the modulation mode identification accuracy of the wireless signal.

[0159] Figure 4 An example of a schematic diagram of a physical structure of an electronic device is shown in FIG. 4. Figure 4 As shown in FIG. 4, the electronic device can include a processor 410, a communications interface 420, a memory 430 and a communications bus 440, wherein the processor 410, the communications interface 420 and the memory 430 can communicate with each other through the communications bus 440. The processor 410 can invoke a logical instruction in the memory 430 to execute a modulation mode identification method based on a comprehensive feature of a wireless signal, the method including: obtaining a comprehensive feature of a target from a wireless signal, the comprehensive feature including at least one of a time domain feature, a frequency domain feature and a time-frequency domain feature; identifying the wireless signal based on a candidate modulation type, and determining a target modulation type as a modulation mode identification result in a case where the comprehensive feature meets a feature decision threshold corresponding to the target modulation type and a signal-to-noise ratio threshold; the candidate modulation type includes at least one of 2ASK, 2 / 4 / 8FSK, MSK, 2 / 4 / 8PSK, OQPSK, pi / 4DQPSK and 16QAM; and the target modulation type belongs to the candidate modulation type.

[0160] In addition, the logic instructions in the memory 430 described above can be implemented in the form of a software function unit and sold or used as an independent product, which can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application essentially or the parts that contribute to the prior art or parts of the technical solutions can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a plurality of instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the method described in various embodiments of the present application. The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes.

[0161] In another aspect, the present application also provides a computer program product, which comprises a computer program, the computer program can be stored on a non-transitory computer readable storage medium, and the computer program is executed by a processor, so that the computer can execute the modulation mode identification method based on the comprehensive features of the wireless signal provided by the above-mentioned method, and the method comprises the following steps: obtaining the comprehensive features of the target from the wireless signal, the comprehensive features comprising at least one of the time domain features, the frequency domain features and the time-frequency domain features; identifying the wireless signal based on the candidate modulation types, and determining the target modulation type as the modulation mode identification result in the case that the comprehensive features meet the feature decision threshold corresponding to the target modulation type and the signal-to-noise ratio threshold; wherein the candidate modulation types comprise at least one of 2ASK, 2 / 4 / 8FSK, MSK, 2 / 4 / 8PSK, OQPSK, π / 4DQPSK and 16QAM; and the target modulation type belongs to the candidate modulation types.

[0162] In another aspect, the present application also provides a computer program product, which comprises a computer program, the computer program can be stored on a non-transitory computer readable storage medium, and the computer program is executed by a processor, so that the computer can execute the modulation mode identification method based on the comprehensive features of the wireless signal provided by the above-mentioned method, and the method comprises the following steps: obtaining the comprehensive features of the target from the wireless signal, the comprehensive features comprising at least one of the time domain features, the frequency domain features and the time-frequency domain features; identifying the wireless signal based on the candidate modulation types, and determining the target modulation type as the modulation mode identification result in the case that the comprehensive features meet the feature decision threshold corresponding to the target modulation type and the signal-to-noise ratio threshold; wherein the candidate modulation types comprise at least one of 2ASK, 2 / 4 / 8FSK, MSK, 2 / 4 / 8PSK, OQPSK, π / 4DQPSK and 16QAM; and the target modulation type belongs to the candidate modulation types.

[0163] The device embodiments described above are merely illustrative, wherein the units described as separate components can or can not be physically separate, and the components displayed as units can or can not be physical units, i.e., can be located in one place, or can be distributed to multiple network units. Part or all of the modules can be selected to achieve the purposes of the embodiments according to actual needs. Those skilled in the art can understand and implement without creative labor.

[0164] Through the description of the above embodiments, those skilled in the art can clearly understand that the embodiments can be realized by means of software and the necessary general hardware platform, and of course can also be realized by hardware. Based on such understanding, the above technical solutions can be embodied in the form of a software product, which can be stored in a computer readable storage medium, such as a ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute the methods described in each embodiment or some parts of the embodiments.

[0165] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A modulation mode recognition method based on comprehensive characteristics of wireless signals, characterized in that, Comprise: Obtaining a comprehensive feature of a target from a wireless signal, the comprehensive feature comprising a time domain feature, a frequency domain feature and a time-frequency domain feature; The comprehensive feature is used to sequentially screen each candidate modulation type in order of the time domain feature, the frequency domain feature and the time-frequency domain feature; Based on the candidate modulation type, the wireless signal is identified, and in the case that the comprehensive feature meets the feature decision threshold and the signal-to-noise ratio threshold corresponding to the target modulation type, the target modulation type is determined as the modulation mode recognition result; wherein the candidate modulation type comprises at least one of 2ASK, 2 / 4 / 8FSK, MSK, 2 / 4 / 8PSK, OQPSK, π / 4DQPSK, CW, 4OQPSK, 2MSK and 16QAM; the target modulation type belongs to the candidate modulation type; the time domain feature comprises envelope smoothness, and the envelope smoothness is used to distinguish amplitude modulation signals, phase modulation signals and frequency modulation signals; The frequency domain feature comprises the peak number and the flatness of the normalized power spectrum, and the frequency domain feature is used to represent the frequency characteristics of the wireless signal; The time-frequency domain feature comprises the flatness of the modulus of the normalized instantaneous frequency, and the time-frequency domain feature is used to identify the state frequency variation characteristics of the wireless signal; The flatness of the modulus of the normalized instantaneous frequency is obtained by the following steps: The instantaneous frequency is obtained by using phase difference, the normalized instantaneous frequency is obtained by normalizing the instantaneous frequency, the modulus of the normalized instantaneous frequency is calculated, and the flatness of the modulus of the normalized instantaneous frequency is evaluated.

2. The modulation mode recognition method based on the comprehensive feature of the wireless signal according to claim 1, wherein The identification of the wireless signal based on the candidate modulation type, in the case that the comprehensive feature meets the feature decision threshold and the signal-to-noise ratio threshold corresponding to the target modulation type, determines the target modulation type as the modulation mode recognition result, comprises: The wireless signal is identified based on the candidate modulation type in the target order, and in the case that the comprehensive feature meets the feature decision threshold and the signal-to-noise ratio threshold corresponding to the target modulation type, the first appearing target modulation type is determined as the modulation mode recognition result; wherein the target order is determined based on the sequence of the time domain feature, the frequency domain feature and the time-frequency domain feature. 3.The method of claim 1, wherein, The time domain feature comprises envelope smoothness; In the case that the comprehensive feature meets the feature decision threshold and the signal-to-noise ratio threshold corresponding to the target modulation type, the target modulation type is determined as the modulation mode recognition result, comprises: In the case that the feature decision threshold corresponding to the time domain feature is greater than 1.27, and the signal-to-noise ratio threshold corresponding to the time domain feature is greater than or equal to 5dB, the modulation mode recognition result is determined as 2ASK; In the case that the feature decision threshold corresponding to the time domain feature is between 1.12 and 1.27, and the signal-to-noise ratio threshold corresponding to the time domain feature is greater than or equal to 12dB, the modulation mode recognition result is determined as 2PSK or 16QAM; in a case where the feature decision threshold corresponding to the time domain feature is between 1.03 and 1.12, and the signal-to-noise ratio threshold corresponding to the time domain feature is greater than or equal to 12 dB, determining that the modulation mode recognition result is 4PSK, 8PSK, pi / 4DQPSK, or 4OQPSK; in a case where the feature decision threshold corresponding to the time domain feature is less than 1.03, and the signal-to-noise ratio threshold corresponding to the time domain feature is greater than or equal to 15 dB, determining that the modulation mode recognition result is CW, 2MSK, or 2 / 4 / 8FSK; or, in a case where the feature decision threshold corresponding to the time domain feature is less than 1.03, and the signal-to-noise ratio threshold corresponding to the time domain feature is greater than or equal to 12 dB, determining that the modulation mode recognition result is 2MSK or 2 / 4 / 8FSK. 4.The method of claim 3, wherein, the frequency domain feature includes a peak value number of a normalized power spectrum; the determining, in a case where the comprehensive feature satisfies a feature decision threshold and a signal-to-noise ratio threshold corresponding to a target modulation type, that the target modulation type is the modulation mode recognition result further includes: in a case where power spectra of the order of 1, 2, 4, and 8 of the wireless signal converge, and the peak value number is 1, determining that the modulation mode recognition result is CW; in a case where power spectra of the order of 1, 2, 4, and 8 of the wireless signal converge, the signal-to-noise ratio threshold corresponding to the power spectrum of the order of 4 is greater than or equal to 0 dB, the signal-to-noise ratio threshold corresponding to the power spectrum of the order of 8 is greater than or equal to 7 dB, and the peak value number is 1, determining that the modulation mode recognition result is 2ASK; in a case where power spectra of the order of 2, 4, and 8 of the wireless signal converge, the signal-to-noise ratio threshold corresponding to the power spectrum of the order of 4 is greater than or equal to 1 dB, the signal-to-noise ratio threshold corresponding to the power spectrum of the order of 8 is greater than or equal to 8.5 dB, and the peak value number is 1, determining that the modulation mode recognition result is 2PSK; in a case where the power spectrum of the order of 4 of the wireless signal converges, the power spectrum of the order of 8 partially converges, the signal-to-noise ratio threshold corresponding to the power spectrum of the order of 4 is greater than or equal to 4 dB, and the peak value number is 1, determining that the modulation mode recognition result is 4PSK or 4OQPSK; in a case where power spectra of the order of 2, 4, and 8 of the wireless signal converge, the signal-to-noise ratio threshold corresponding to the power spectrum of the order of 2 is greater than or equal to 1 dB, the signal-to-noise ratio threshold corresponding to the power spectrum of the order of 4 is greater than or equal to 7 dB, the signal-to-noise ratio threshold corresponding to the power spectrum of the order of 8 is greater than or equal to 13.5 dB, and the peak value number is 2, determining that the modulation mode recognition result is 2MSK. 5.The method of claim 4, wherein, the frequency domain feature further includes a flatness degree; the determining, in a case where the comprehensive feature satisfies a feature decision threshold and a signal-to-noise ratio threshold corresponding to a target modulation type, that the target modulation type is the modulation mode recognition result further includes: In a case where the characteristic decision threshold of the power spectrum of the 1st power of the wireless signal is greater than 0.3, the corresponding signal-to-noise ratio threshold is greater than -6 dB, the characteristic decision threshold of the power spectrum of the 2nd power is greater than 0.32, the corresponding signal-to-noise ratio threshold is greater than 1 dB, the characteristic decision threshold of the power spectrum of the 4th power is greater than 0.27, and the corresponding signal-to-noise ratio threshold is greater than 3 dB, the modulation mode recognition result is determined to be 2ASK. In a case where the characteristic decision threshold of the power spectrum of the 1st power of the wireless signal is less than 0.3, the corresponding signal-to-noise ratio threshold is greater than -6 dB, the characteristic decision threshold of the power spectrum of the 2nd power is greater than 0.32, the corresponding signal-to-noise ratio threshold is greater than 1 dB, the characteristic decision threshold of the power spectrum of the 4th power is greater than 0.27, and the corresponding signal-to-noise ratio threshold is greater than 3 dB, the modulation mode recognition result is determined to be 2PSK. In a case where the characteristic decision threshold of the power spectrum of the 1st power of the wireless signal is less than 0.3, the corresponding signal-to-noise ratio threshold is greater than -6 dB, the characteristic decision threshold of the power spectrum of the 2nd power is less than 0.1, the corresponding signal-to-noise ratio threshold is greater than or equal to -6 dB, and the power spectrum of the 4th power satisfies one of the following conditions: In a case where the characteristic decision threshold of the power spectrum of the 4th power is between 0.1 and 0.4, and the corresponding signal-to-noise ratio threshold is greater than 3 dB, the modulation mode recognition result is determined to be 4PSK or π / 4DQPSK. In a case where the characteristic decision threshold of the power spectrum of the 4th power is between 0.25 and 0.4, and the corresponding signal-to-noise ratio threshold is greater than 6 dB, the modulation mode recognition result is determined to be 4OQPSK. In a case where the characteristic decision threshold of the power spectrum of the 4th power is less than 0.05, and the corresponding signal-to-noise ratio threshold is greater than or equal to -6 dB, the modulation mode recognition result is determined to be 8PSK. In a case where the characteristic decision threshold of the power spectrum of the 1st power of the wireless signal is less than 0.3, the corresponding signal-to-noise ratio threshold is greater than -6 dB, the characteristic decision threshold of the power spectrum of the 2nd power is less than 0.3, the corresponding signal-to-noise ratio threshold is greater than or equal to -3 dB, and the power spectrum of the 4th power satisfies one of the following conditions: In a case where the characteristic decision threshold of the power spectrum of the 4th power is less than 0.05, and the corresponding signal-to-noise ratio threshold is greater than -6 dB, the modulation mode recognition result is determined to be 8FSK.

6. The modulation mode identification method based on comprehensive characteristics of wireless signals according to claim 4 or 5, characterized in that, The time-frequency domain feature includes a flatness of a modulus of a normalized instantaneous frequency. In a case where the comprehensive feature satisfies a characteristic decision threshold and a signal-to-noise ratio threshold corresponding to a target modulation type, the target modulation type is determined to be the modulation mode recognition result, which further includes: In a case where the flatness of the modulus of the normalized instantaneous frequency corresponds to a characteristic decision threshold greater than 1.95, and the corresponding signal-to-noise ratio threshold is greater than or equal to 8 dB, the modulation mode recognition result is determined to be 2ASK, 2 / 4 / 8PSK, or 16QAM. In a case where the flatness of the modulus of the normalized instantaneous frequency corresponds to a characteristic decision threshold between 1.32 and 1.95, and the corresponding signal-to-noise ratio threshold is greater than or equal to 12 dB, the modulation mode recognition result is determined to be 4OQPSK or 4 / 8FSK. In a case that the flatness degree of the module corresponds to a feature decision threshold between 1.05 and 1.32, and a corresponding signal-to-noise ratio threshold is greater than or equal to 20 dB, the modulation mode identification result is determined as 2FSK or 2MSK; In a case that the flatness degree of the module corresponds to a feature decision threshold less than 1.05, and a corresponding signal-to-noise ratio threshold is greater than or equal to 13 dB, the modulation mode identification result is determined as CW.

7. A device for identifying modulation modes based on comprehensive features of wireless signals, applying the method for identifying modulation modes based on comprehensive features of wireless signals according to claim 1, characterized in that, The method comprises: a feature acquisition module, configured to acquire a comprehensive feature from a wireless signal, the comprehensive feature comprising a time domain feature, a frequency domain feature, and a time-frequency domain feature; the comprehensive feature is used to sequentially screen each candidate modulation type in order of the time domain feature, the frequency domain feature, and the time-frequency domain feature; an identification module, configured to identify the wireless signal based on a candidate modulation type, and in a case that the comprehensive feature satisfies a feature decision threshold and a signal-to-noise ratio threshold corresponding to a target modulation type, determine the target modulation type as a modulation mode identification result; wherein the candidate modulation type comprises at least one of 2ASK, 2 / 4 / 8FSK, MSK, 2 / 4 / 8PSK, OQPSK, π / 4DQPSK, and 16QAM; and the target modulation type belongs to the candidate modulation type.

8. An electronic device comprising a memory, a processor, and a computer program stored on the memory and running on the processor, characterized in that, The processor executes the computer program to implement the modulation mode identification method based on a comprehensive feature of a wireless signal according to any one of claims 1 to 6. 9.A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the modulation mode identification method based on a comprehensive feature of a wireless signal according to any one of claims 1 to 6.

Images